Efficiency differentials in peasant agriculture and their implications for development policies

This paper reviews the theoretical and empirical work on allocative efficiency in traditional agriculture and presents a new study of technical efficiency among Tanzanian cotton farmers. The theoretical arguments are shown to apply primarily in a competitive context that differs significantly from that in which peasant farms operate. Re‐analysis of earlier empirical studies shows that, on average, the marginal value products of inputs differ by more than 40 per cent from the marginal factor costs to which they should be equated under allocative efficiency. Our own study among Tanzanian cotton farmers in Geita District reveals that output could be increased by 51 per cent if all farmers achieved those levels of technical efficiency that were in fact achieved by the best farmers in the sample using the same inputs and technologies that the less efficient used. These results indicate that the efficiency hypothesis may not be applicable to much of peasant agriculture and that development policies might fruitfully place more emphasis on raising large numbers of farmers closer to the relatively high efficiency levels achieved by some of their neighbors.